Skip to content

spranab/brainstorm-mcp

v1.5.6 Breaking

This release includes 1 breaking change for platform teams planning a safe upgrade.

Published 1mo MCP Developer Tools
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

ai ai-debate brainstorming claude deepseek groq
+7 more
llm mcp mcp-server model-context-protocol multi-model ollama openai

Summary

AI summary

Dropped the temperature field for all gpt-5.x reasoning models, which reject non-default values.

Full changelog

Fix

Drop temperature field for gpt-5.x reasoning models.

The useNewTokenParam branch already correctly used max_completion_tokens for these models, but it still passed temperature: 0.7 which all gpt-5.x reasoning models reject with:

400 Unsupported value: 'temperature' does not support 0.7 with this model. Only the default (1) value is supported.

This affected gpt-5, gpt-5.1, gpt-5.2, gpt-5.3, gpt-5.4, gpt-5.5 (released 2026-04-23), and the o-series.

What still works

The older gpt-4.x branch (useNewTokenParam=false) keeps temperature: 0.7 since gpt-4.x models still accept it.

Verified

  • Direct probe of gpt-5.5-2026-04-23: HTTP 200, valid completion, no temperature in request body
  • Hit while running a 3-round brainstorm with openai:gpt-5.5 + deepseek:deepseek-chat — gpt-5.5 voice errored every round-1 turn until this fix.

Breaking Changes

  • Removed `temperature` field for all gpt-5.x reasoning models (e.g., gpt-5, gpt-5.1‑5.5, o-series).

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track spranab/brainstorm-mcp

Get notified when new releases ship.

Sign up free

About spranab/brainstorm-mcp

Multi-round AI brainstorming debates between multiple models (GPT, Gemini, DeepSeek, Groq, Ollama, etc.). Pit different LLMs against each other to explore ideas from diverse perspectives.

All releases →

Beta — feedback welcome: [email protected]